🔍 Read the full analysis: Keep AI Financial Help Within Safe Boundaries on ThorstenMeyerAI.com
Get the latest gadgets delivered free with Prime
- Fast, free delivery on millions of items
- Prime Video, Amazon Music and more included
- Member-only deals all year
TL;DR
Bank of America has warned customers not to give banking usernames or passwords to AI tools, saying customers may be responsible for activity resulting from shared access. The warning comes as financial firms discuss AI-led payments and trading, but the legal treatment of losses caused by an agent remains unsettled.
Bank of America has warned customers not to share their banking usernames or passwords with AI tools or agents, saying they are responsible for activity resulting from that access. The warning arrives as financial companies discuss AI agents making payments and trades, leaving customers to weigh convenience against potential account and liability risks.
The email, titled “Protecting your finances in the AI World,” says credentials should never be shared with another person or provided to an outside service, including an AI-enabled tool or agent acting on a customer’s behalf. Bank of America also said it is “excited about what AI used responsibly can deliver.” A customer posted the email on Bluesky; TechCrunch reported it on October 27, and a bank spokesperson confirmed its authenticity to the publication.
The bank’s warning is about a basic problem with password-based access: an agent using a customer’s login can appear to the bank as the customer. That can make it harder to distinguish an action the customer wanted from one caused by an error, malicious instruction or unauthorized use. Banks may also block activity they consider suspicious, potentially interrupting the customer’s access.
On October 28, the source reported that Visa had told investors trust was the biggest barrier to AI agents making payments. Robinhood, meanwhile, said customers had put more than $250 million into accounts that AI agents trade. Those examples indicate that agent activity is already part of the financial-services conversation, but they do not establish how often agents are making transactions or what protections apply across different products.
Before your agent touches money
Give it a valet key, not the master key
Logging in with your password makes the agent you. A credential made for the agent can do only the job you gave it, and you can take it back.
- What it opensEverything you can do: transfers, payees, settings, statements
- LimitsNone that the agent cannot get around by logging in as you
- What the bank seesYou, or a thief. It cannot tell which
- If a web page tricks itThe trick reaches your whole account
- To take it backChange your password everywhere it was used
- If money goes missingUS rules can exclude transfers by someone you gave access to; Bank of America: “you’re responsible”
- What it opensOnly what you allowed: read-only, one card, one pot of money
- LimitsBuilt in: amount, merchant, time window, approval on your phone
- What the bank seesAn identified agent acting for an identified person
- If a web page tricks itThe damage stops at the limit
- To take it backDisconnect the agent or delete its card. Your password never moved
- If money goes missingDepends on the terms. Read them first: some still put all risk on you
A comparison of two set-ups, not a measurement. Which limits exist depends on your bank, card and app.
Sources: Bank of America customer email, confirmed to TechCrunch (27 Oct 2026); 12 CFR 1005.2(m); OWASP LLM01. Comparison: Thorsten Meyer.
Powered by Thorsten Meyer AI
The Risk of Giving Agents Account Access
Sharing a password gives an agent broad access, rather than permission for one limited task. If an agent can log in as the customer, it may be able to take actions beyond the user’s immediate request, depending on the account’s features and the bank’s controls. The same access can also make an agent’s actions difficult to separate from a customer’s own activity.
The financial consequences are not settled simply by adding an approval screen or a spending cap. In the United States, Regulation E defines an unauthorized electronic fund transfer in a way that excludes certain transfers made by someone the customer furnished an access device to. The source’s author argues that this language could matter when a customer gives credentials to an agent or its operator, but says no regulator or court has ruled on that application. It should not be treated as a definitive legal conclusion.
In the European Union, the Payment Services Directive requires payment-service users to take reasonable steps to keep personalized security credentials safe. The source notes that whether sharing a password with an AI agent would count as gross negligence under the directive has not been tested in court. For consumers, the practical point is that technical safeguards and legal responsibility are separate questions: a product may limit an agent’s actions, while its terms still assign risk to the customer.
Products Worth Considering
Samsung 27 Odyssey OLED G5 (G50SF) QHD QD-OLED Gaming Monitor
LG UltraGear 32GX850A-B 32-Inch 4K UHD Glossy OLED Gaming Monitor
Samsung 27-inch Odyssey OLED G6 QD-OLED QHD Gaming Monitor
As an affiliate, we earn on qualifying purchases.
Banks Confront Agents in Accounts
The warning follows a report by Axios on October 22, which cited unnamed sources saying major U.S. banks were finding third-party AI agents inside customer accounts, often logged in with the customers’ own passwords. According to that report, banks were struggling to tell whether the activity reflected the customer’s intent. These claims are attributed to Axios’s sources; the material provided does not independently verify how widespread the activity is.
The security concern is not limited to a password being stolen. AI agents can read webpages, emails, invoices and files, and text in those materials may attempt to direct an agent to take actions the user did not request. This is commonly called prompt injection. The source says OWASP ranks it as a leading risk for AI applications. The specific likelihood or financial impact of such an attack on any individual bank account is not established in the material.
The source recommends narrower forms of permission instead of sharing a full login: read-only access, requiring approval for each payment, a virtual card with a set limit, or a token restricted to a particular merchant, amount and time period. A separate account or pool of funds can also limit potential losses. These are proposed safeguards, not guarantees that an agent will behave correctly or that a provider will accept responsibility for losses.
“Your username and password “should never be shared with another person or provided to an outside service, including an AI-enabled tool or agent that offers to act on your behalf.””
— Bank of America, in its customer email
Liability Rules Remain Untested
No court or regulator ruling is identified that decides how existing U.S. or EU payment rules apply when a customer shares credentials with an AI agent. The exact outcome could depend on the service, the way access was granted, the transaction and the applicable law. The source’s discussion of legal exposure is an interpretation, not legal advice.
It is also unclear how many customers are using agents to access bank accounts, how frequently banks are blocking those agents, and what safeguards are in place across providers. The Robinhood figure describes money placed in accounts that AI agents trade; it does not say that agents have generated returns, losses or a particular volume of trades. The source does not provide details about the controls or terms attached to every agent-payment product mentioned.
Check Permissions Before Connecting
Customers considering an AI financial tool should read its permissions and terms before linking an account, and avoid giving it a bank password where a narrower access method is available. They should check whether transactions require approval, whether spending limits can be set, and what the provider says about responsibility for errors or losses. A separate account with limited funds may reduce exposure, but does not remove all risk.
For financial firms and agent developers, the immediate issue is how to offer task-specific access without relying on customers’ full credentials, and how to explain who bears responsibility when an agent makes a disputed transaction. The material provides no confirmed timetable for new rules or product changes. Banks’ responses, future regulatory guidance and any court decisions will help clarify what protections customers can expect.
Key Questions
What did Bank of America tell customers?
It said customers should not share banking usernames or passwords with another person or an outside service, including an AI agent. The email says customers are responsible for activity resulting from shared access.
Does sharing credentials automatically mean a customer must pay for losses?
That is not established by the source. Bank of America states its position in the email, while the application of U.S. and EU payment rules to AI-agent access has not been resolved in a court ruling identified here.
What is a safer way to let an AI agent handle a payment?
The source recommends limiting permission—for example, read-only access, payment-by-payment approval, a capped virtual card, or a token limited by merchant, amount and time. Availability and protections vary by provider.
What does the Robinhood figure mean?
Robinhood said customers had put more than $250 million into accounts that AI agents trade. The source gives no comparison baseline or measurement period, and the figure does not establish trading results or losses.
Source: ThorstenMeyerAI.com
